prof.&romanmartoňák department*of*experimental*physics,* *comenius… · 2012-07-24 ·...

1
Simulations of structural phase transitions in crystals by metadynamics Prof. Roman Martoňák, Department of Experimental Physics, Comenius University, Mlynská dolina Bratislava, Slovakia Thursday July 26, 11:00, Room MXC 315 Abstract: Crystal structure prediction problem is an important problem in physics, chemistry and materials science. Besides methods based on stochastic generation of candidate structures a possible approach is to study structural phase transitions between different polymorphs. In addition to information about new structures the latter approach provides also useful hints to microscopic mechanisms of structural transitions on atomistic level which are difficult to extract from experiments. Computer simulations of structural transformations in crystals, however, present a non trivial problem due to high barriers separating different crystalline phases. The talk will present an overview of basic principles and some applications of the metadynamicsbased algorithm for simulation of structural phase transitions in crystals [1,2,3,4]. The approach is based on the general metadynamics algorithm [5] and uses the supercell matrix as collective variable driving the system from one crystal structure to another, following low freeenergy pathways. By effectively introducing a mechanism allowing crossing of barriers it eliminates the need for substantial overpressurization typically needed in constantpressure MD simulations, resulting in more realistic transformation mechanisms as well as substantially improved predictive power. It also enables simulation of complex reconstructive transitions with transformation pathways proceeding via several intermediate states. The algorithm has been successfully applied to study of highpressure structural transformations in various materials, such as MgSiO3 (post perovskite phase) [6], SiO2 [7], Si [8], CO2 [9], C [10], Ca [11], BN [12] etc. References : [1] R. Martoňák, A. Laio, and M. Parrinello: Phys. Rev. Lett. 90, 075503 (2003) [2] R. Martoňák, D. Donadio, A. R. Oganov and M. Parrinello: Nature Materials 5, 623 (2006) [3] R. Martoňák, Eur. Phys. J. B 79, 241 (2011) [4] R. Martoňák, chapter in Modern Methods of Crystal Structure Prediction, edited by A. R. Oganov (WileyVCH, Berlin, 2011) [5] A. Laio and M. Parrinello, PNAS 99, 12562 (2002) [6] A. R. Oganov, R. Martoňák, A. Laio, P. Raiteri and M. Parrinello, Nature 438, 1142 (2005) [7] D. Donadio, R. Martoňák, P. Raiteri and M. Parrinello: Phys. Rev. Lett. 100, 165502 (2008) [8] J. Behler, R. Martoňák, D. Donadio and M. Parrinello: Phys. Rev. Lett. 100, 185501 (2008) [9] J. Sun, D. D. Klug, R. Martoňák, J. A. Montoya, M.S. Lee, S. Scandolo and E. Tosatti, PNAS 106, 6077 (2009) [10] J. Sun, D. D. Klug, R. Martoňák, J. Chem. Phys. 130, 194512 (2009) [11] Y. Yao, D. D. Klug, J. Sun and R. Martoňák, Phys. Rev. Lett. 103, 055503 (2009) [12] L. Hromadová and R. Martoňák, Phys. Rev. B 84, 224108 (2011)

Upload: vuphuc

Post on 15-Feb-2019

214 views

Category:

Documents


0 download

TRANSCRIPT

                                   

Simulations  of  structural  phase  transitions  in  crystals  by  metadynamics  

 Prof.  Roman  Martoňák,  Department  of  Experimental  Physics,    Comenius  University,  Mlynská  dolina  Bratislava,  Slovakia  

Thursday  July  26,  11:00,  Room  MXC  315    

       Abstract:  Crystal  structure  prediction  problem  is  an  important  problem  in  physics,  chemistry  and  materials  science.  Besides  methods  based  on  stochastic  generation  of   candidate   structures  a  possible  approach   is   to   study   structural  phase  transitions  between  different  polymorphs.  In  addition  to  information  about  new  structures  the  latter  approach  provides  also  useful  hints  to  microscopic  mechanisms  of  structural  transitions  on  atomistic  level  which  are  difficult  to  extract   from  experiments.  Computer   simulations  of   structural   transformations   in  crystals,  however,  present  a  non-­‐trivial  problem  due  to  high  barriers  separating  different  crystalline  phases.  The  talk  will  present  an  overview  of  basic  principles  and  some  applications  of  the  metadynamics-­‐based  algorithm  for  simulation  of  structural  phase  transitions  in  crystals  [1,2,3,4].  The  approach  is  based  on  the  general  metadynamics  algorithm  [5]  and  uses  the  supercell  matrix  as  collective  variable  driving  the  system  from  one  crystal  structure  to  another,  following  low  free-­‐energy  pathways.  By   effectively   introducing   a   mechanism   allowing   crossing   of   barriers   it   eliminates   the   need   for   substantial  overpressurization   typically  needed   in  constant-­‐pressure  MD  simulations,   resulting   in  more  realistic   transformation  mechanisms  as  well  as  substantially  improved  predictive  power.  It  also  enables  simulation  of  complex  reconstructive  transitions   with   transformation   pathways   proceeding   via   several   intermediate   states.   The   algorithm   has   been  successfully  applied  to  study  of  high-­‐pressure  structural  transformations  in  various  materials,  such  as  MgSiO3  (post-­‐perovskite  phase)  [6],  SiO2  [7],  Si  [8],  CO2  [9],  C  [10],  Ca  [11],  BN  [12]  etc.        References  :    [1]  R.  Martoňák,  A.  Laio,  and  M.  Parrinello:  Phys.  Rev.  Lett.  90,  075503  (2003)  [2]  R.  Martoňák,  D.  Donadio,  A.  R.  Oganov  and  M.  Parrinello:  Nature  Materials  5,  623  (2006)  [3]  R.  Martoňák,  Eur.  Phys.  J.  B  79,  241  (2011)  [4]   R.  Martoňák,   chapter   in  Modern  Methods   of   Crystal   Structure   Prediction,   edited   by   A.   R.   Oganov   (Wiley-­‐VCH,  Berlin,  2011)  [5]  A.  Laio  and  M.  Parrinello,  PNAS  99,  12562  (2002)  [6]  A.  R.  Oganov,  R.  Martoňák,  A.  Laio,  P.  Raiteri  and  M.  Parrinello,  Nature  438,  1142  (2005)  [7]  D.  Donadio,  R.  Martoňák,  P.  Raiteri  and  M.  Parrinello:  Phys.  Rev.  Lett.  100,  165502  (2008)  [8]  J.  Behler,  R.  Martoňák,  D.  Donadio  and  M.  Parrinello:  Phys.  Rev.  Lett.  100,  185501  (2008)  [9]  J.  Sun,  D.  D.  Klug,  R.  Martoňák,  J.  A.  Montoya,  M.-­‐S.  Lee,  S.  Scandolo  and  E.  Tosatti,  PNAS  106,  6077  (2009)  [10]  J.  Sun,  D.  D.  Klug,  R.  Martoňák,  J.  Chem.  Phys.  130,  194512  (2009)  [11]  Y.  Yao,  D.  D.  Klug,  J.  Sun  and  R.  Martoňák,  Phys.  Rev.  Lett.  103,  055503  (2009)  [12]  L.  Hromadová  and  R.  Martoňák,  Phys.  Rev.  B  84,  224108  (2011)